Llama 4 Scout vs Qwen3.8 27B
Updated October 21, 2026. 58 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Pick Qwen3.8 27B for coding: it scores 73.6 on the Vibe Coding Index against 9.1 for Llama 4 Scout.
Pick Llama 4 Scout when output tokens cost 4.5x less, or when the context window is larger, 1.31M against 1M tokens.
- Shared results
- 58
- Llama 4 Scout only
- 22
- Qwen3.8 27B only
- 27
- Like-for-like categories
- 7 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, four public coding benchmarks
Qwen3.8 27B
Qwen3.8 27B scores 73.6 against 9.1.
1 and 2 benchmarks behind the scores
Intelligence
AA Intelligence Index
Qwen3.8 27B
Qwen3.8 27B leads, 33.7 against 6.5.
Same board, both measured
Agentic work
AA Agentic Index
Qwen3.8 27B
Qwen3.8 27B leads, 45.8 against 0.5.
Same board, both measured
Reasoning
GPQA Diamond
Qwen3.8 27B
Qwen3.8 27B leads, 88.9% against 47.0%.
Same board, both measured
Math
AIME
No call
Only Llama 4 Scout has a AIME result, 19.0%.
Directional only
Knowledge
MMLU Pro
Qwen3.8 27B
Qwen3.8 27B leads, 84.3% against 69.6%.
Same board, both measured
Web design
Design Arena: websites
No call
Only Llama 4 Scout has a Design Arena: websites result, 750.
Directional only
Long documents
Listed context window
Llama 4 Scout
Llama 4 Scout takes 1.31M tokens against 1M.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Llama 4 Scout
Llama 4 Scout $0.0003, Qwen3.8 27B $0.0007 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
Qwen3.8 27B
Llama 4 Scout $1.276, Qwen3.8 27B $1.26.
Published attempt cost and pass rate
At a glance
- Coding index
- Llama 4 Scout 9.1 from 1 benchmark, Qwen3.8 27B 73.6 from 2 benchmarks.
- Widest category
- Agentic work, 45.3 points apart on average over 1 shared benchmark.
- Widest single gap
- ARC-AGI: Llama 4 Scout 0.5% against Qwen3.8 27B 87.5%, 87.0 points apart on a 0 to 100 scale.
- Output price
- Llama 4 Scout $0.30 against Qwen3.8 27B $1.35 per million output tokens, 4.5x apart.
- Context window
- Llama 4 Scout takes 1.31M tokens, Qwen3.8 27B 1M.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Llama 4 Scout | Bars | Qwen3.8 27B | Ties |
|---|---|---|---|---|
| Coding | 0 | 5 | ||
| Intelligence | 0 | 2 | ||
| Agentic work | 0 | 1 | ||
| Reasoning | 0 | 11 | ||
| Math | 0 | 2 | ||
| Knowledge | 0 | 9 | ||
| General | 0 | 28 | ||
| All shared boards | 0 | 58 |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index05
| Benchmark | Llama 4 Scout | Qwen3.8 27B | Leader |
|---|---|---|---|
| SWE-bench Verified | 9.1% | Not measured | Only Llama 4 Scout measured |
| LiveBench coding | Not measured | 86.8% | Only Qwen3.8 27B measured |
| LMArena WebDev | Not measured | 1593 | Only Qwen3.8 27B measured |
| AA Coding Index | 8.2 | 68.1 | Qwen3.8 27B |
| LiveCodeBench (Vals AI) | 38.5% | 84.0% | Qwen3.8 27B |
| SWE-bench Verified (Vals AI) | Not measured | 86.0% | Only Qwen3.8 27B measured |
| Vibe Code Bench (Vals AI) | Not measured | 64.8% | Only Qwen3.8 27B measured |
| Code migration (Vals AI) | Not measured | 14.2% | Only Qwen3.8 27B measured |
| ProgramBench (Vals AI) | Not measured | 0.0% | Only Qwen3.8 27B measured |
| IOI (Vals AI) | Not measured | 39.1% | Only Qwen3.8 27B measured |
| Terminal-Bench 2.1 (Vals AI) | Not measured | 58.4% | Only Qwen3.8 27B measured |
| SkillsBench (Vals AI) | Not measured | 38.1% | Only Qwen3.8 27B measured |
| CyberBench (Vals AI) | Not measured | 67.8% | Only Qwen3.8 27B measured |
| LiveCodeBench (Artificial Analysis run) | 29.9% | Not measured | Only Llama 4 Scout measured |
| SciCode | 21.3% | 46.6% | Qwen3.8 27B |
| Terminal-Bench Hard (Artificial Analysis run) | 1.5% | Not measured | Only Llama 4 Scout measured |
| LMArena Text: coding | 1286 | 1483 | Qwen3.8 27B |
| LMArena Text: software and IT | 1286 | 1478 | Qwen3.8 27B |
| LMArena WebDev: React | Not measured | 1602 | Only Qwen3.8 27B measured |
| LMArena WebDev: HTML | Not measured | 1570 | Only Qwen3.8 27B measured |
| LMArena WebDev: from an image | Not measured | 1577 | Only Qwen3.8 27B measured |
| Neither model has a result on the other 17 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Llama 4 Scout | Qwen3.8 27B | Leader |
|---|---|---|---|
| AA Intelligence Index | 6.5 | 33.7 | Qwen3.8 27B |
| Epoch Capabilities Index | 129.6 | 151.7 | Qwen3.8 27B |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Llama 4 Scout | Qwen3.8 27B | Leader |
|---|---|---|---|
| AA Agentic Index | 0.5 | 45.8 | Qwen3.8 27B |
| LiveBench research | Not measured | 59.4% | Only Qwen3.8 27B measured |
| LiveBench browser | Not measured | 32.0% | Only Qwen3.8 27B measured |
| LiveBench long agentic | Not measured | 27.9% | Only Qwen3.8 27B measured |
| Finance Agent (Vals AI) | Not measured | 48.5% | Only Qwen3.8 27B measured |
| Tax Agent Bench (Vals AI) | Not measured | 30.9% | Only Qwen3.8 27B measured |
| Legal Research Bench (Vals AI) | Not measured | 36.1% | Only Qwen3.8 27B measured |
| Excel Modeling Benchmark (Vals AI) | Not measured | 59.7% | Only Qwen3.8 27B measured |
| Terminal-Bench-Science (Vals AI) | Not measured | 0.0% | Only Qwen3.8 27B measured |
| τ²-Bench Telecom | 15.5% | Not measured | Only Llama 4 Scout measured |
| APEX-Agents | Not measured | 47.5% | Only Qwen3.8 27B measured |
| Neither model has a result on the other 28 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index011
| Benchmark | Llama 4 Scout | Qwen3.8 27B | Leader |
|---|---|---|---|
| GPQA Diamond | 47.0% | 88.9% | Qwen3.8 27B |
| LiveBench reasoning | Not measured | 80.0% | Only Qwen3.8 27B measured |
| MMMU Pro (Vals AI) | 58.8% | 83.9% | Qwen3.8 27B |
| GPQA Diamond (Artificial Analysis run) | 58.7% | 90.5% | Qwen3.8 27B |
| Humanity's Last Exam (Artificial Analysis run) | 3.8% | 33.9% | Qwen3.8 27B |
| GPQA Diamond (Epoch AI run) | 51.8% | Not measured | Only Llama 4 Scout measured |
| ARC-AGI | 0.5% | 87.5% | Qwen3.8 27B |
| ARC-AGI-2 | 0.0% | 42.4% | Qwen3.8 27B |
| LMArena Text: hard prompts | 1266 | 1459 | Qwen3.8 27B |
| LMArena Text: expert | 1235 | 1479 | Qwen3.8 27B |
| LMArena Text: hard prompts in English | 1293 | 1461 | Qwen3.8 27B |
| LMArena Vision: diagrams | 1125 | 1285 | Qwen3.8 27B |
| LMArena Vision: homework | 1148 | 1295 | Qwen3.8 27B |
| Neither model has a result on the other 8 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Llama 4 Scout | Qwen3.8 27B | Leader |
|---|---|---|---|
| AIME | 19.0% | Not measured | Only Llama 4 Scout measured |
| MATH 500 (Vals AI) | 79.2% | Not measured | Only Llama 4 Scout measured |
| MGSM (Vals AI) | 88.0% | Not measured | Only Llama 4 Scout measured |
| AA Math Index | 14.0 | Not measured | Only Llama 4 Scout measured |
| AIME 2025 (Artificial Analysis run) | 14.0% | Not measured | Only Llama 4 Scout measured |
| OTIS Mock AIME | 7.8% | Not measured | Only Llama 4 Scout measured |
| ProofBench | Not measured | 16.0% | Only Qwen3.8 27B measured |
| MATH level 5 | 62.3% | Not measured | Only Llama 4 Scout measured |
| LMArena Text: math | 1287 | 1455 | Qwen3.8 27B |
| LMArena Text: mathematical work | 1286 | 1458 | Qwen3.8 27B |
| Neither model has a result on the other 3 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index09
| Benchmark | Llama 4 Scout | Qwen3.8 27B | Leader |
|---|---|---|---|
| MMLU Pro | 69.6% | 84.3% | Qwen3.8 27B |
| LegalBench (Vals AI) | 72.0% | 82.4% | Qwen3.8 27B |
| CorpFin (Vals AI) | 46.8% | Not measured | Only Llama 4 Scout measured |
| TaxEval (Vals AI) | 55.2% | 70.8% | Qwen3.8 27B |
| MortgageTax (Vals AI) | 57.8% | 64.9% | Qwen3.8 27B |
| MedQA (Vals AI) | 50.9% | Not measured | Only Llama 4 Scout measured |
| MedCode (Vals AI) | 23.3% | 28.7% | Qwen3.8 27B |
| MMLU Pro (Artificial Analysis run) | 75.2% | Not measured | Only Llama 4 Scout measured |
| LMArena Text: science | 1278 | 1467 | Qwen3.8 27B |
| LMArena Text: medicine and healthcare | 1275 | 1437 | Qwen3.8 27B |
| LMArena Text: legal and government | 1285 | 1447 | Qwen3.8 27B |
| LMArena Text: business and finance | 1257 | 1438 | Qwen3.8 27B |
| Neither model has a result on the other 5 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
| Benchmark | Llama 4 Scout | Qwen3.8 27B | Leader |
|---|---|---|---|
| Design Arena: websites | 750 | Not measured | Only Llama 4 Scout measured |
| Design Arena: UI components | 767 | Not measured | Only Llama 4 Scout measured |
| Design Arena: game dev | 773 | Not measured | Only Llama 4 Scout measured |
| Design Arena: data viz | 896 | Not measured | Only Llama 4 Scout measured |
| Design Arena: all code categories | 790 | Not measured | Only Llama 4 Scout measured |
| LiveBench design | Not measured | 75.9% | Only Qwen3.8 27B measured |
| Neither model has a result on the other 3 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index028
| Benchmark | Llama 4 Scout | Qwen3.8 27B | Leader |
|---|---|---|---|
| LiveBench data analysis | Not measured | 76.6% | Only Qwen3.8 27B measured |
| SAGE (Vals AI) | 34.8% | 52.4% | Qwen3.8 27B |
| MedScribe (Vals AI) | 50.6% | 83.8% | Qwen3.8 27B |
| IFBench | 39.5% | Not measured | Only Llama 4 Scout measured |
| Long context reasoning (AA-LCR) | 27.7% | 82.0% | Qwen3.8 27B |
| DTBench | 57.9% | 88.0% | Qwen3.8 27B |
| LMCA | 12.0% | 41.4% | Qwen3.8 27B |
| Fiction.LiveBench | 36.0% | Not measured | Only Llama 4 Scout measured |
| Surface Evolver Bench | Not measured | 45.0% | Only Qwen3.8 27B measured |
| LMArena Text | 1279 | 1440 | Qwen3.8 27B |
| LMArena Text: instruction following | 1248 | 1436 | Qwen3.8 27B |
| LMArena Text: creative writing | 1249 | 1383 | Qwen3.8 27B |
| LMArena Vision | 1118 | 1269 | Qwen3.8 27B |
| LMArena Text: English | 1304 | 1448 | Qwen3.8 27B |
| LMArena Text: non-English | 1252 | 1428 | Qwen3.8 27B |
| LMArena Text: Chinese | 1255 | 1501 | Qwen3.8 27B |
| LMArena Text: German | 1273 | 1443 | Qwen3.8 27B |
| LMArena Text: Japanese | 1206 | 1387 | Qwen3.8 27B |
| LMArena Text: Korean | 1207 | 1384 | Qwen3.8 27B |
| LMArena Text: French | 1283 | 1465 | Qwen3.8 27B |
| LMArena Text: Spanish | 1278 | 1449 | Qwen3.8 27B |
| LMArena Text: Russian | 1263 | 1415 | Qwen3.8 27B |
| LMArena Text: Polish | 1265 | 1439 | Qwen3.8 27B |
| LMArena Text: multi-turn | 1280 | 1440 | Qwen3.8 27B |
| LMArena Text: longer queries | 1265 | 1448 | Qwen3.8 27B |
| LMArena Text: writing and language | 1269 | 1405 | Qwen3.8 27B |
| LMArena Text: entertainment and media | 1242 | 1393 | Qwen3.8 27B |
| LMArena Vision: English | 1143 | 1267 | Qwen3.8 27B |
| LMArena Vision: Chinese | 1106 | 1326 | Qwen3.8 27B |
| LMArena Vision: reading text | 1117 | 1281 | Qwen3.8 27B |
| LMArena Vision: humor | 1102 | 1235 | Qwen3.8 27B |
| LMArena Vision: creative writing | 1126 | 1237 | Qwen3.8 27B |
| Neither model has a result on the other 14 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Llama 4 Scout | Qwen3.8 27B | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Llama 4 Scout costs less
Repository review
50K input + 3K output tokens
Llama 4 Scout costs less
Agent loop
200K input + 10K output tokens
Qwen3.8 27B costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Meta Llama 4 Scout | Qwen3.8 27B |
|---|---|---|
| Maker | Meta Llama | Qwen |
| Weights | Open weights | Open weights |
| Released | April 5, 2025 | August 14, 2026 |
| Context window | 1.31M | 1M |
| Input $/Mtok | $0.10 | $0.04 |
| Output $/Mtok | $0.30 | $1.35 |
| Coding index | 9.1 | 73.6 |
| Independent measurements behind it | 1 | 2 |
| Cost per solved task | $1.276 | $1.26 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Llama 4 Scout better than Qwen3.8 27B for coding?
Pick Qwen3.8 27B for coding: it scores 73.6 on the Vibe Coding Index against 9.1 for Llama 4 Scout. Pick Llama 4 Scout when output tokens cost 4.5x less, or when the context window is larger, 1.31M against 1M tokens.
Which is cheaper, Llama 4 Scout or Qwen3.8 27B?
Llama 4 Scout costs $0.10 per million input tokens and $0.30 per million output tokens; Qwen3.8 27B costs $0.04 and $1.35. Per solved SWE-bench Verified task, Llama 4 Scout costs $1.276 and Qwen3.8 27B $1.26.
Which has the bigger context window, Llama 4 Scout or Qwen3.8 27B?
Llama 4 Scout, with 1.31M tokens against 1M.
How many benchmarks is the Llama 4 Scout vs Qwen3.8 27B comparison based on?
58 benchmarks have a result for both models as of October 21, 2026. Llama 4 Scout's coding index rests on 1 independent benchmark and Qwen3.8 27B's on 2. Where only one model has a result, the page says so instead of naming a winner.